Gamma Scalping

Owning an option is really a bet on movement. You don't need the stock to go up or down, you need it to move a lot. Gamma scalping is the machinery that turns that movement into cash. The recipe: buy an option, strip out its directional bet by hedging, and then let the hedge itself do the earning.

Here's the intuition. When you own a call and the stock rises, your position automatically becomes "longer", it acts like you own more and more shares. To stay direction-neutral you must sell some shares. When the stock falls, your position becomes "shorter," so you buy shares back. Sell as it rises, buy as it falls: you are mechanically buying low and selling high, over and over, and pocketing the little round-trips. That automatic re-hedging is gamma scalping, and the reason it keeps handing you profitable trades is gamma, the curvature of the option's value.

Where the profit comes from

For a delta-hedged option position, the profit and loss over a small time step is captured by one clean relationship:

$$\text{P\&L} \approx \tfrac{1}{2}\,\Gamma\,(\Delta S)^2 \;+\; \Theta\,\Delta t.$$

In words: $\Gamma$ (gamma) is how sharply your directional exposure bends as the stock moves; $\Delta S$ is the size of the stock's move; $\Theta$ (theta) is the value your option bleeds as time passes; and $\Delta t$ is the time elapsed. The first term is your gamma gain, and notice the $(\Delta S)^2$: it doesn't care whether the stock went up or down, only how far. The second term is your theta bill, and it's negative when you own options. Big moves feed you; the passing of time starves you.

Worked example

You own a call with gamma $\Gamma = 0.02$ per point on a $100 stock, and it costs you theta of $4 per day. You delta-hedge with shares so you have no direction. Over one day the stock swings, and you re-hedge on the way, netting a round-trip roughly equivalent to one clean $2 move ($\Delta S = 2$):

• Gamma gain: $\tfrac{1}{2}\,\Gamma\,(\Delta S)^2 = \tfrac{1}{2}(0.02)(2^2) = \tfrac{1}{2}(0.02)(4) = 0.04$ per share of exposure, which on a decent position scales up to real money, say $5 for the day.
• Theta bill: you owe $4 for the day just for holding the option.
• Net: $5 - 4 = +1$. The stock moved enough to more than cover its rent.

Now suppose the stock barely twitches, a $1-equivalent move. The gamma gain falls to $\tfrac{1}{2}(0.02)(1) = 0.01$ per share, maybe $1.25 for the day, but you still owe the $4 theta. Net: a loss of $2.75. Same position, same option, opposite result, decided entirely by how much the stock moved.

The break-even is realized versus implied

That last example is the entire game. There is a level of movement at which the gamma gains exactly cover the theta bill, and it is not arbitrary: it's the implied volatility baked into the option's price. You paid for the option at some implied vol; that price is the market charging you a theta that matches the movement it expects.

• If the stock realizes more volatility than implied, your scalps beat your theta, gamma scalping makes money.
• If it realizes less, theta wins, and you bleed.

So a long-gamma, delta-hedged book is a pure bet: realized volatility above implied. This is the mechanical heart of Volatility Arbitrage.

What eats the profit

• Transaction costs. Every scalp crosses a bid-ask spread and pays fees. Re-hedge too often and costs devour the theoretical gains; too rarely and you leave scalps on the table and carry directional risk. The optimal hedge frequency trades one off against the other.
• Gaps. The formula assumes small, continuous moves. A jump, an earnings surprise, a halt, delivers a move your hedge never caught, and the tidy $(\Delta S)^2$ math breaks.
• The other side is brutal. Selling options and scalping the short gamma flips every sign: you collect theta but lose on movement, small steady gains punctuated by rare, violent losses. Short gamma is how desks blow up.

The bookkeeping behind the P&L formula lives in Delta-Hedging P&L, the Greeks it uses in The Option Greeks, and the rent you're paying in Theta and Time Decay.